mechanism of polysome function, involving ... - Europe PMC

600

BIOCHEMISTRY: WILSON AND HOAGLAND

PROC. N. A. S.

* This investigation was supported by a research grant from the U.S. Public Health Service (AI-04479). 1 Aposhian, H. V., and A. Kornberg, J. Biol. Chem., 237, 519 (1962). 2 Lucas, Z. J., and A. Kornberg, personal communications. 3Lucas, Z. J., Federation Proc., 24, 286 (1965). Oleson, A. E., and J. F. Koerner, J. Biol. Chem., 239, 2935 (1964). Richardson, C. C., C. L. Schildkraut, H. V. Aposhian, and A. Kornberg, J. Biol. Chem., 239,

222 (1964).

6Mokrasch, L. C., and R. W. McGilvery, J. Biol. Chem., 221, 909 (1956). Layne, E., in Methods in Enzymology, ed. S. P. Colowick and N. 0. Kaplan (New York: Academic Press, Inc., 1957), vol. 3, p. 451. 8 Koerner, J. F., M. S. Smith, and J. M. Buchanan, J. Biol. Chem., 235, 2691 (1960). 9 Shortman, K., and I. R. Lehman, J. Biol. Chem., 239, 2964 (1964). 10 Jorgensen, S. E., and J. F. Koerner, Federation Proc., 24, 349 (1965). '1 Jorgensen, S. E., unpublished experiments. I

STUDIES ON THE PHYSIOLOGY OF RAT LIVER POLYRIBOSOMES: QUANTITATION AND INTRACELLULAR DISTRIBUTION OF RIBOSOMES* BY SAMUEL H. WILSON AND 1\IAHLON B. HOAGLAND DEPARTMENT OF BACTERIOLOGY AND IMMUNOLOGY, HARVARD MEDICAL SCHOOL, BOSTON, MASSACHUSETTS

Communicated by Paul Doty, May 27, 1965

The central role of polyribosomes in cytoplasmic protein synthesis has been demonstrated in a wide variety of organisms. In the rat liver system, polysomes, when isolated, are aggregates of 83S ribosomes joined together, at least in part, by material sensitive to ribonuclease and resistant to trypsin and deoxyribonuclease, presumably messenger RNA. Consistent with presently favored hypotheses of the mechanism of polysome function, involving reattachment to messenger RNA of monosomes from a relatively large cytoplasmic pool,1' 2 investigators have observed a high proportion of "free" (70-85S) ribosomes in preparations derived from many sources. As techniques for polysome isolation have improved, however, the proportion of 70-85S ribosomes has diminished.3 4 Early in the course of our investigation of rat liver polysome physiology, it became apparent that 260-m/i absorption of sucrose gradient fractions is not an accurate indication of the quantity of polysomal RNA. Indeed, most of the 260mju optical density in the 70-85S region is due to the metalloprotein, ferritin.5 On correcting for ferritin, the proportion of deoxycholate-liberated monosomes is reduced to less than 5 per cent of total cytoplasmic ribosomal RNA. Evidence will be presented that no additional free and/or deoxycholate-liberated cytoplasmic monosomes are being degraded, compartmentalized, or otherwise selected against during isolation l)rocedures and sucrose gradient analysis. Those free ribosomes that do exist in liver extracts are primarily confined to the nucleus. Materials and Methods.-6-C14-Orotic acid hydrate, 5 mc/mmole, was obtained from New England Nuclear Corporation and dissolved in dilute NaOH before intraperitoneal injection. Sodium deoxycholate and sodium dodecyl sulfate (from Mann Research Laboratories, and Merck,

VOL. 54, 1965


601

Sharpe and Dohme, respectively) were freshly prepared in water before each procedure. Female Sprague-Dawley rats weighing 200 gm were obtained from the Charles River Breeding Laboratories, Wilmington, Mass. Trypsin, RNase, and DNase were obtained from the Worthington Biochemical Corporation. Fractionation of liver: Normal, well-fed rats were killed by decapitation and their livers were quickly excised and chilled. All subsequent procedures were carried out at 0-20C. The livers were weighed in medium X (0.15 M sucrose, 0.025 M KCl, 0.008 M MgCl2, 0.1 M Tris buffer, pH 7.4 at 20C), minced for 15 sec, and homogenized in 2 vol of medium X with three quick strokes in a loose-fitting Potter-Elvehjem homogenizer at 400 rpm. Unbroken cells, nuclei, mitochondria, and miscellaneous cell debris were removed by centrifugation at 15,000 g for 4.5 min in a Servall RC-2 centrifuge using a model SS-34 rotor.6 Under these conditions of homogenization, about half the microsomal content of the cells was liberated into the supernatant (as determined by glucose-6-phosphatase assay: see Table 1), and all the nuclei in the pellet appeared intact upon microscopic examination. Isolation of ribosomal material from 15,000 g supernatant: Polysomes were isolated by a modification of the zone centrifugation technique of Wettstein et al.7 The 15,000 g supernatant was quickly decanted and made 1.3% with respect to DOC. After gentle stirring for 1 min the deep red, opaque mixture was layered over 10 ml of medium X containing 0.5 M sucrose, which had previously been layered over 10 ml of medium X containing 1.0 M sucrose in a 36-ml tube for the Spineo no. 30 rotor. Procedures up to this point usually required 15-20 min. The material was then centrifuged at 78,500 g average for 8 hr at 0C in a Spinco model L2 preparative ultracentrifuge. The polysome pellet was resuspended in 2 ml of medium X by gentle homogenization (leaving behind the underlying clear glycogen pellet) and subjected to sucrose gradient analysis. Other methods of isolation failed to yield as high a proportion of polysomal particles (>128S). Temperature elevation during fractionation and centrifugation proved to be the single most important factor contributing to the apparent breakdown of heavy polysomal material. By direct assay;8 polysomes were found to be essentially free of DOC. The nonferritin protein to RNA ratio was consistently 0.8-1.0, but tended to vary greatly with the quantity of DOC added to the 15,000 g supernatant. Thus, a final DOC concentration of 0.5% gave a nonferritin protein to RNA ratio of 1.7, while a final concentration of 2% gave a ratio of 0.70-0.85. Isolation of ribosomal material from the 15,000 g pellet: A 15,000 g pellet derived from 7 gm of liver, prepared as described in the preceding section, was taken up in 14 ml of medium X, thoroughly homogenized, and made 2.0% with respect to DOC. After gentle stirring for 1 min the viscous material was subjected to the zone centrifugation described above. The resulting pellet was taken up in 8 ml of medium X and analyzed by sucrose gradient centrifugation. DNase treatment of this pellet produced no change in sucrose gradient oD profiles. As will be shown, nuclei in the 15,000 g pellet were a rich source of free ribosomes and disomes. The latter were probably magnesium-dependent, sedimenting at 128S,9 since they were resistant to ribonuclease treatment. For the isolation of radioactive free ribosomes, animals were injected with 10 uc C14-orotic acid and the ribosomal material of the 15,000 g pellet was isolated, as described above, 4 hr later. Sucrose gradients were then run, and the monosome and disome fractions were pooled, diluted with cold medium X lacking sucrose, and resedimented in the Spinco no. 40 rotor by centrifugation at 40,000 rpm for 6 hr. The pellets were washed by resuspension and recentrifugation and submitted once again to sucrose gradient analysis, after which the ribosome fractions were TABLE 1 QUANTITATION OF HOMO(GENATE COMPONENTS

Homogenate 15O0 g SN 15,O00 9 Pellet*

RNA

G-6-Pase

Ferritin

ribosomes

100 16 ± 6 84 ± 6

100 43 ± 4 59 ± 5 Mg/gm wvet wt 9.0 ±(0.5

100 56 : a5 42 ± 5

100 70 ± 5 30 ± 5

100 29 ± 4.5 70 ± 4.5

Mg/gm wret wvt

Homogenate *

Per cent total

DNA

3.4 i 0.6

The assay on the 15,000 g pellet was performed after thorough homogenization.

602


PROC. N. A. S.

again sedimented as usual and stored at -70'C in pellet form. This procedure obviated the use of ribonuclease in preparing monosomes from polysomes and ensured relative freedom from ferritin. Isolation of ribosomes from purified nuclei: Free nuclei were isolated according to Maggio et al.,'0 except that centrifugation was carried out for 55 instead of 90 min, to minimize contamination with microsomal material. As demonstrated by phase contrast microscopy and glucose-6phosphatase assays, the nuclei were intact and essentially free of whole cells, cell debris, or microsomal material. According to DNA estimations done on the homogenate and nuclear pellet, about 4% of total liver nuclei was recovered. Ribosomal material was extracted from these nuclei exactly as described in the preceding section. Isolation of ferritin and purified RNA: An entire polysome pellet derived from one rat liver was taken up in 2 ml of a solution of 0.5% SDS, 0.01 M Tris, and 0.1 M sucrose. The suspension was incubated at 370C for 4 min, then 1 ml was layered directly over a 25-ml, 15-35% linear sucrose density gradient containing 0.1 M Tris. Centrifugation was carried out for 18 hr at 58,000 g using an SW25 rotor in a Spinco model L ultracentrifuge. The fractions containing ferritin were visually identified, pooled, diluted with 0.01 M Tris, and centrifuged at 105,000 g for 4 hr. The resulting reddish-brown ferritin pellet was dissolved in 2 ml 0.01 M Tris, treated with 30 ,g RNase at 370C for 30 min, and resedimented by a 4-hr centrifugation at 105,000 g. The ferritin was taken up once again in 0.01 M Tris, dialyzed against water, and subjected to sucrose gradient analysis, whereupon it yielded a single symmetrical peak. Ferritin prepared in this manner was free of RNA according to the estimation procedures cited below. When isolation of RNA was desired, sucrose gradients as described above were monitored for optical density in a continuous-flow Gilford spectrophotometer, and fractions containing RNA were collected in a chilled centrifuge tube, diluted, and sedimented at 105,000 g for 7 hr. The pellets were then taken up in a small amount of 0.5% SDS and extracted with phenol for 15 min at 60C with vigorous shaking. The aqueous layer was removed and made 0.1 M with respect to NaCl and 67% with respect to ethanol, and placed at -150C for 16 hr to facilitate precipitation. Sucrose density gradient centrifugations: Linear gradients of 15-35% sucrose in medium X were prepared according to the technique of Britten and Roberts." Quantities up to 2.5 mg RNA were layered over such gradients. Polysome material was usually centrifuged for 90 min at 25,000 rpm at 00C. Following centrifugation, fractions were collected through a Sigma Co. finger-type pump at a rate which usually delivered 0.96 ml to each of 26 chilled collection tubes every 30 sec. In this manner, fluctuation in drop size leading to fractionation error was min-

imized. When resuspended polysome pellets were analyzed according to this technique, 30-45% of total RNA layered on the gradient was found at the bottom of the tube and therefore not monitored for OD. This material was included when quantitating polysomes. Ultraviolet absorption analysis: Chilled, diluted sucrose gradient fractions were adjusted to the approximate concentration of optimum Zeiss spectrophotometer reading efficiency, and optical density at both 260 and 320 m1A was recorded. The surprisingly high absorption due to sucrose and other gradient components was corrected for by reading the OD of the low- and high-concentration sucrose solutions and plotting a straight line between the two values over the range of 26 fractions. The corresponding contaminant absorption was subtracted from each reading of the experimental sample. Chemical analyses and measurement of radioactivity: RNA, DNA, and protein were estimated by standard procedures.'2-14 Polysomal RNA was estimated directly using the extinction coefficient of 26.5 OD units/(mg)(ml)(cm). Glucose-6-phosphatase was used as a marker for endoplasmic reticular membranes of the cytoplasm and was assayed according to the technique of De Duve."5 Aliquots of diluted sucrose gradient fractions were counted for 20 min in a Nuclear-Chicago scintillation counter at a counting efficiency for C14 RNA of 72%. All measurements were corrected for background and quenching.

Results.- Correction of sucrose gradients for ferritin content: Ferritin and 83128S particles from the 15,000 g pellet were purified as described under Methods.


VOL. 54, 1965

603

The absorption spectra are presented in Figure 1. 12.2 It is seen that ferritin has a relatively high extine. tion at 260 mjh. A correction for ferritin is readily effected by read'4 .4 1.0ing the sample at a wavelength, such as 320 mA, at 0.2?9 which ribosomal RNA does not absorb, but which is 220 240 260 280 300 320 within the range of the spectrophotometer's hydrogen me lamp effectiveness. By reference to the ferritin spec- traFIG.of 1.-UV spec( ferritin absorption *)and trum it is seen that absorption is 1.59 times higher at purified ribosomes ---- -0). 260 mjA than at 320 m~i. Thus, the reading of the ex- Both materials were dissolved perimental sample at 260 mjg minus 1.59 times its reading at 320 mgu is the 260-miA absorption due to nonferritin material. The validity of this correction was confirmed by two further experiments. (a) Sucrose gradient analysis of radioactive free ribosomes and disomes, prepared as described in Methods, allowed determination of their specific radioactivity (cpm/OD, Fig. 2). Labeled ribosomes and ferritin were then mixed and subjected to gradient centrifugation. The OD profile obtained by applying the ferritin correction was essentially identical to that obtained by correcting for the dilution of specific radioactivity due to ferritin. (b) Direct RNA analysis on each fraction20 yielded gradient profiles identical to those obtained by applying the ferritin correction. When the ferritin correction is applied to the absorption of isolated polysomes and of a 15,000 g supernatant fraction, a striking change occurred: the OD peak in the 83S region was virtually eliminated (Fig. 3). Intracellular distribution of ribosomes: The apparent paucity of free ribosomes in the 15,000 g supernatant fraction is by no means proof that a large cytoplasmic "pool" of nonpolysomal monosomes does not exist, since a surprisingly small fraction (6-8%) of total liver RNA is monitored in our polysomal sucrose gradients. It was pointed out in the Methods section that 30-45 per cent of ribosomal material put on our gradients sediments to the bottom of the tube during centrifugation. Further, as seen in Figure 4, a large proportion (59%) of RNA is found in the 15,000 g pellet, of which about one third is ribosomal. Thus, about 31 per cent (11 plus 20) of total liver RNA can be accounted for as ribosomal. It has already been emphasized that polysomal material isolated from the 15,000 g supernatant fraction contains very few monosomes (Figs. 3 and 4). Centrifugation of this fraction at 78,000 g for longer periods, up to 12 hr, yielded no more ribosomal

1.0

E 0.6

cu

0

I So

600

R~o^ / 1\

ICbc'

0.4

400

d

0.2 5

IO TUBE NUMBER

200 15

>

FIG. 2.-Sucrose density gradient of a mixture of ferritin and purified radioactive monosomes and disomes. Centrifugation was carried out at 25,000 rpm for 5 hr. Curve a (0 *) is total OD / b (o- - -- -0) is total radio\\measured, activity measured, and c (A----- A) is true ribosomal OD calculated by dividing the total radioactivity of each fraction by the specific radioactivity of the ribosomes (980 cpm/OD). The same curve is obtained by applying our ferritin correction to the observed OD profile. Curve A) is ferritin as determined by d (t the ferritin correction.


604

PROC. N. A. S.

1.6 A

of (A) a 15,000 g superFIG. 3.-Sucrose density gradients natant fraction before DOC treatment and (B) a usual polysome preparation. Two mg of RNA were layered over each gradient. Centrifugation was carried out at 25,000 rpm at 0C for 4 hr in (A) and 1.5 hr in (B). Since materials other than ferritin contributed to the OD320 in (A), the ferritin OD320 was determined according to the following formula: (nonferritin 0D260/nonferritin OD320) (total OD320 - x) + 1.59 * x = total OD260, where x = ferritin OD320. (The nonferritin OD ratio was obtained by measuring the ratio in the heavier fractions of the for the ferritin correction applied gradient.) The isformula total to (A) and (B) as follows: OD260- (ferritinas 01)320 1.59) = polysomal OD260 0- 0.

1.2

E2.8 \

/

oS

o 0.4 a

V

*-.-.-.A

*.2-

B

-

1285 1>265.4S ~~~~~~ ~ 128as ~~~~~~~~~~~~~~